What really is Business Intelligence one may ask? I had noted earlier in my blog about the ‘Convergence of SOA and BI’ that Business Intelligence enlightens an organization about Information (and it has to be relative to the business process). I went on to note that Statistics (why things are happening), Trends and Forecasting / Predictive Modeling (what will happen next and will it continue) and finally Business process optimization (what’s the best that can occur) all require an understanding of the data in context.
Data in context says a lot – let me start by saying that we are in the information age and data is everywhere and information is not! Let me also start with an assertion that business needs information to thrive. Data must be integrated, cleansed, refreshed and turned into reliable business information in order for a corporation to flourish. Information is as critical to a corporations’ success today as water is to life on this planet and without a clean supply of integrated information, corporations will labor to even exist!
The value of information is also directly proportional to how fast a business can react to it – in other words, a corporation loses money every time it delays getting information into the hands of decision makers. Nowhere was this as clearly demonstrated than in a graph created by Dr. Richard Hackathorn called the “Time-Value Curve”. If you never saw it, it shows that the value of data is highest when the event occurs and then it slopes downward. The delay in time it takes for a user to see it and then to react to it he gave metrics such as ‘analysis latency’ and ‘decision latency’. The time it took to take action was the ‘value lost!’
Wikipedia defines BI as ‘“Business intelligence (BI) refers to skills, knowledge, technologies, applications, quality, risks, security issues and practices used to help a business to acquire a better understanding of market behavior and commercial context.’ What it fails to mention first is the most important – data is not information. Information is made up of data but it is data with understanding and a single dictionary (one dictionary breeds understanding, two or more breed confusion).
Data is raw ‘junk’ and consists of both structured (db’s, files) and non-structured components (Documents, Web pages, etc.). Structured data is data with corresponding metadata – or data with its syntax but t is still not information yet. Semantics give meaning to structured data elements but this is not information. Information is structured data combined with semantics (or what the data is in the context of the business. Knowledge goes on to create purposeful combinations of information – giving the information relevance. Dr. Peter Aiken has been speaking about this for years and it seems that people are just now starting to understand what he has been raving about.
So, raw data reporting is the display of data with a simple description of the field. Performance reporting, the first phase of BI, is placing information (businesses facts in context with semantics) in the context of your business model and operations process. Data Warehousing, the more mature phase of BI, is making sense of these facts so that we can now derive conclusions from related and non-related information. Business Performance Management [BPM], an advanced form of BI, is making coherent judgments and inferences from the knowledge gained by evaluating all the possible outcomes and with the understanding of patterns (from the D/W phase) gives ‘benefits driven usage’ or options.
Most in this field have seen the Data Triangle – Data at the bottom, Information above this, knowledge above that and Wisdom at the top of the triangle. This is where we all are trying to get to! The elusive peak of information is wisdom!
I will next describe how different business stakeholders require different types of information!
Thanks for this post. For novices in BI like me, it is a tremendous value add. I would say a couple of points:
1. Time value curve: I work extensively on Oracle Applicatios in HRMS and Purchasing modules. I have some exposure in allied Business Intelligence modules. The idea of these modules is to automate data into information based on some rules and setups. I believe such rules are intended to reduce latencies. However, with dynamic nature of business, sometimes the information siphoned due to these setups cannot be used in as-is leading to the need of human intervention and thereby injecting latency. How best we can address this aspect?
2. You comment on wisdom is really interesting. I wonder do we need DW to be wise in our day to day life. We all know plastics are not good for the environment and yet all our models might tell a computer packaging company to use certain kind of plastic in certain quantities to save on costs. In effect, decision made is wise or not is hard to debate.
Thanks Rama
"Business Intelligence enlightens an organization about Information (and it has to be relative to the business process)".
I suggest that too much of so called BI actually floods the business managers with information about what happened, when and where. At best that should be seen as "knowledge" i.e. data in context.
Intelligence is the process of identifying and highlighting the exceptions that warrant attention i.e. require, or will benefit from, action.